π0.5: A VLA with open-world generalization

Re:π0.5: A VLA with open-world generalization by beklein on Tuesday April 22, 2025 @02:06PM (#43764764)

This is amazing! As someone working with industrial robots, normally under strict environmental constraints and control, witnessing such real-world robotics progress truly excites me about the future!

By the way, they’ve open-sourced their π0 model (code and model weights). More information can be found here: https://github.com/Physical-Intelligence/openpi

[ Parent | on HN ]

Re:π0.5: A VLA with open-world generalization by UltraSane on Tuesday April 22, 2025 @02:17PM (#43764851)

It seems robotics has advanced more in the last 3 years than the previous 20.

[ Parent | on HN ]

Re:π0.5: A VLA with open-world generalization by htrp on Tuesday April 22, 2025 @08:30PM (#43767585)

the torrent of funding helps here

[ Parent | on HN ]

Re:π0.5: A VLA with open-world generalization by Tireings on Wednesday April 23, 2025 @02:02AM (#43769080)

ML helps here and the progress Nvidia made with their robotics platform.

[ Parent | on HN ]

Re:π0.5: A VLA with open-world generalization by rurban on Wednesday April 23, 2025 @07:53AM (#43771040)

But mostly OpenCV, in its excellent C++ and python variants. Not everything is modern ML heuristics, some classic AI is also needed still.

[ Parent | on HN ]

Re:π0.5: A VLA with open-world generalization by UltraSane on Wednesday April 23, 2025 @06:40AM (#43770540)

The vision language action models and the two level slow planning and fast control LLMs seem to be a big breakthrough.

[ Parent | on HN ]